JNTU B.Tech III DATA ANALYTICS Previous Question Paper (March-2021 - R18)

A hazard is any source, situation, or act with the potential to cause harm, injury, or adverse health effects to individuals, or damage to property and the environment. In an organizational context, hazards represent risks that can disrupt operations or compromise safety.

Potential sources of hazards include:

• Physical Hazards: These include noise, radiation, extreme temperatures, vibration, and poor lighting, which can lead to immediate or long-term physical injury.
• Chemical Hazards: Exposure to hazardous substances like solvents, cleaning agents, dust, or fumes that can cause respiratory issues, skin irritation, or systemic toxicity.
• Biological Hazards: Exposure to bacteria, viruses, fungi, or other organic materials, particularly relevant in healthcare or laboratory settings.
• Ergonomic Hazards: Factors such as repetitive movements, improper workstation design, or heavy lifting that lead to musculoskeletal disorders.
• Psychosocial Hazards: Workplace stress, bullying, excessive workload, or lack of role clarity, which impact mental health and productivity.
• Safety Hazards: Mechanical or electrical risks, such as unguarded machinery, faulty wiring, or slippery floors, which pose an immediate threat of accidents.

Model-based techniques are analytical approaches that rely on the creation of a mathematical or logical representation (a model) of a system or process to predict behavior, identify patterns, or optimize outcomes. Unlike data-driven approaches that focus purely on statistical correlation, model-based techniques incorporate domain knowledge and structural assumptions about the underlying system.

Key characteristics and examples include:

• Structural Representation: These techniques define relationships between variables based on theoretical understanding, such as physical laws or business logic.
• Simulation: Models are used to simulate various scenarios to observe how the system responds to different inputs, which is useful when real-world experimentation is costly or dangerous.
• Optimization: Techniques like Linear Programming or Markov Decision Processes use models to find the best possible solution under specific constraints.
• Predictive Accuracy: By formalizing the system's mechanics, model-based techniques often provide better interpretability and can function with smaller datasets compared to purely black-box machine learning models.